In-Depth Analysis of Fatal Motorcycle Accidents—Case Study in Slovenia

Abstract

Motorcyclists remain a disproportionately large group of vulnerable road users, with fatality rates significantly higher than that in other road groups. Additionally, fatal accidents involving motorcyclists have a more slowly decreasing trend in comparison to that of other road users, while the number of this kind of users is growing fast. For all these reasons, there is a need to understand what the key factors leading to fatal accidents are in order to identify the possible measures to minimize the accidents themselves or at least their consequences. This would permit, indeed, to positively impact the road traffic system, leading to the creation of the safest road traffic system possible, as it is the goal of the Sustainable Safety approach. The aim of this study is to dive into the mentioned problem, analyzing fatal motorcycle accidents in Slovenia over a decade, highlighting the key factors contributing to these incidents. By integrating data from four databases, the study evaluated accident trends, infrastructural elements, and rider behavior through a multi-stage analysis. Firstly, data were collected from four national, up-to-date databases that contain information about road accidents themselves, the road infrastructure, additional police data, and media descriptions. After merging this information into one comprehensive database, where each row represents all the data available for one accident, a general analysis of accidents’ trends over the considered 10-year period was developed, considering at first all fatal road accidents, then deepening it to accidents caused by a motorcyclist, and finally to single-vehicle accidents. A statistical analysis followed, aimed at identifying a statistical correlation between the accidents and the factors leading to them. The results of the first accident analysis indicated that excessive speed, incorrect driving direction, and overtaking maneuvers are the primary causes of fatal accidents, especially on non-urban roads preferred by motorcyclists. Single-vehicle accidents frequently involve collisions with roadside objects, including safety barriers and poles, underscoring the need for targeted infrastructural improvements. The following correlation analysis revealed that a total of seven factors were statistically significant: three human factors (age, gender, experience)—which were the ones with the strongest correlations—one infrastructural factor (pavement width), and three factors belonging to external conditions (accident type, cause, and location). Of these, four were positively correlated to the causer, while three, i.e., pavement width, causes, and road location, were negatively correlated. This study provides a foundation for future research on less severe accidents and proactive risk behavior analysis, aiming to improve motorcyclist safety comprehensively.

1. Introduction

It is well known that motorcyclists are vulnerable road users. This is due to many aspects related to both the vehicle and the rider, such as the different sizes, weights, and powers of the motorcycle, the balance skills needed by the rider to control the bike, the high speeds that motorcycles can reach, and the low level of protection of their users, who are only protected by their safety equipment—if it is used effectively [1].

Motorcycle designs vary significantly in size, weight, and performance capacity, allowing riders to choose models tailored to specific riding practices [2]. However, the likelihood of motorcyclists being killed in road accidents remains disproportionately high—9 to 30 times greater than that of car drivers [3]. The situation is further complicated by the global increase in motorcycle ownership, which contributes to changing road safety dynamics and patterns of risk exposure [4].

Although the number of motorcyclist fatalities has decreased over the last ten years [5], the decrease has been much slower than that of other road users: a 25% decrease compared to 33% of all other road fatalities. In other words, over 45,000 motorcyclists were killed on European roads over the last 10 years. At the same time, it should be noted that the number of motorcyclists is increasing, and the patterns of road deaths are changing [5]. Concurrently, the growing number of motorcyclists has introduced new challenges in maintaining road safety. These trends underscore the urgency of deepening our understanding of the factors contributing to fatal motorcycle accidents. Although technological advancements such as enhanced braking systems can mitigate some risks, the persistent over-representation of motorcyclists in fatality statistics highlights the need for broader and more targeted interventions.

Motorcycle safety predominantly affects men. Of all motorcycle-related fatalities, 92% involve male riders, 2% male passengers, 3% female riders, and 3% female passengers. For example, the corresponding numbers for Slovenia are 93%, 1%, 0%, and 6%. In countries tracking speed compliance on urban roads with a 50 km/h limit, 41% to 61% of powered two-wheeler (PTW) speed observations in free-flowing traffic adhere to the legal limit. For rural non-motorway roads, compliance ranges from 27% to 81% of PTW speed observations. On motorways, compliance with speed limits for PTWs in free-flowing traffic varies between 42% and 85% [5]. For all these reasons, it is necessary to deepen the knowledge of the factors that lead to fatal accidents involving motorcyclists: only in this way is it possible to propose appropriate countermeasures that could reduce the number or at least the consequences of road accidents involving these users.

The present study aims to shed light on the safety level of motorcyclists in Slovenia by analyzing the most recent road accident databases available at the national level to determine which factors influence the occurrence of fatal accidents caused by motorcyclists.

2. Literature Review

For several reasons, the motorcycle–rider–environment system is more demanding than the driver–vehicle–environment system. Firstly, the elements of motorcycle construction differ from that of passenger cars. Secondly, rider behavior differs from driver behavior [6], also between urban and rural roadways [7]. Thirdly, some road design elements (or their combinations) and road equipment elements could be completely safe for motorized vehicles but can be very dangerous for motorcyclists (and there is no difference in terms of the road or roadside elements). For example, curves present hazardous sections for motorcyclists. Motorcycles require specific riding skills; riders can quickly lose control over the motorcycle (steering and navigating through curves) [8]. Additionally, motorcyclist’s roadway perception and visibility are degraded in horizontal curves [9], which is especially evident on rural roads [10].

The issue of motorcyclist safety started getting the attention of researchers in the early 2000s when Clarke et al. [11] were among the first researchers to conduct an in-depth analysis of motorcyclist road accidents. They used both the analysis of a police database of road accidents involving motorcyclists and the results of questionnaires to identify the most common types of accidents involving these road users and the leading causes. Three causes of accidents were highlighted: accidents due to failure to give way, accidents due to losing control of the motorcycle, and overtaking. A few years later, Clabaux et al. [12] pointed out another important aspect of motorcyclists’ safety: their conspicuity. According to these researchers, many accidents are caused by interfering users who do not notice the motorcyclist despite seeing them. They have shown that one of the main factors influencing the conspicuity of motorcyclists is their speed: the higher their speed, the less visible they are to other road users.

Recently, two interesting studies have been conducted, one in the United States [13] and the other in the Czech Republic [14], in which various aspects of motorcyclist safety were investigated. In the first study, Jafari et al. [13] focused on road infrastructure. They emphasized that curved elements are more challenging for motorcyclists due to vehicle handling and balancing.

In the second study, Tmejova et al. [14] examined 376 accidents involving motorcyclists in South Moravia. According to their results, due to incorrect speed, incorrect assessment of the situation, and inexperience, motorcyclists caused half of the road accidents. As confirmed by Clarke et al. [11], rider inexperience is a consistent risk factor globally, especially among young or newly licensed motorcyclists. An interesting study was recently published by Bella et al. [15], who focused on the overtaking maneuver, which has its peculiarities, especially for motorcyclists. They developed an experiment in which a car equipped with a camera and a GPS device recorded the overtaking maneuvers of motorcyclists. Based on this experiment, they analyzed the duration of overtaking maneuvers and the lateral clearance distance maintained by motorcyclists, factors that can significantly affect safety when overtaking. Recent work by Promraksa et al. [16] studying the factors affecting lateral clearance confirms its importance during filtering maneuvers.

Although the efforts towards technical improvements of motorcyclist safety in Slovenia are remarkable, little scientific work has been carried out on these road users. One paper should be pointed out: Šraml et al. [17] prepared an analysis of the road safety of PTWs on Slovenian territory in 2012 and showed the main causes leading to road accidents involving PTWs. However, recent advancements in data collection and accident analysis methodologies offer new opportunities for a deeper understanding of these road users’ safety [18].

The following paper presents a recent analysis of road accidents that happened on Slovenian roads involving motorcyclists. The paper is structured as follows: first, a literature review was conducted to show the already-known aspects of fatal accidents caused by motorcyclists. Secondly, the databases and methods used to obtain valuable data on fatal motorcycle accidents are described. Finally, four analyses were conducted: first, an analysis of road accidents over ten years from the entire database considered—considering only motorcyclist accidents—was made to show recurrent patterns in the evolution of fatal accidents involving the mentioned road users; this analysis was then deepened by considering only accidents caused by motorcyclists to understand what factors influenced the misbehavior of these road users. Finally, of all the accidents caused by a motorcyclist, only single-vehicle accidents were considered. In the end, a correlation analysis was carried out to understand which underlined factors also statistically influence the occurrence of such accidents.

3. Methodology

Four complementary databases were used as the starting point of this study to develop accurate, up-to-date, and comprehensive research. These are:

• The road accident database by the Slovenian Traffic Safety Agency (TSA) contains essential information on road accidents recorded by the police from 1994 to 2022. In this database, the data on road accidents are recorded for each road accident separately and contain information on the time and location of the road accident, its typology, cause, classification, consequences, the characteristics of the road which the accident occurred on, as well as much information about the users involved in the accidents. The data are displayed interactively on a road accident map on the website of TSA and are open to be downloaded as an .xlsx file.
• The road data bank (RDB) is a record of technical data about public—national and municipal—roads, facilities, traffic signs, and additional equipment present on these roads. This database is maintained and upgraded by the Slovenian Infrastructure Agency DRSI, which takes care of data about national roads, and by municipalities for the data about municipal roads.
• Additional police data were used to obtain additional detailed information not contained in the TSA database. These data describe events that happened on Slovenian roads in the last 24 h. Records are available by days, months, and years and contain brief statistical overviews of what happened in the last 24 h, reports of stolen vehicles, and descriptions of fatal road accidents.
• Additional online media descriptions often report additional details about fatal road accidents that happened on Slovenian roads.

To develop the desired analyses, it was necessary to create a unique database containing data on the road accidents themselves, data on the geometrical characteristics of the road, and the road users involved. To do so, four steps were followed (Figure 1). Firstly, valuable data were extracted from TSA and RDB databases (Figure 1—step 1). From the TSA database, data from 1 January 2013–31 December 2022 were chosen to analyze a 10-year period—this timeframe was considered because it is a long enough period to obtain a sufficient quantity of data and is recent enough to ensure an up-to-date analysis.

We should point out that motorways and expressways were excluded from the analysis, because in the structure of the traffic flow on Slovenian motorways and expressways, there is a negligible share of motorcyclists. Differently, on other national (state) roads, the presence of motorcyclists is significantly higher. As a result, accidents involving motorcyclists on motorways and expressways are rare exceptions.

Among the reasons for this different use of the roads by motorcyclists, there is the monotonous riding (no curves) on motorways and expressways, the large number of trucks on these roads (which are generally prohibited on other national roads), the motorways and expressways tolls, and the unsightly surroundings of motorways. Additionally, the speed limit on Slovenian motorways is 130 km/h (on national roads outside settlements it is generally 90 km/h), which riders of certain types of motorcycles (custom, roadster) do not want to drive. In the TSA database, we used two additional filters to get the data for our study: the participant (of the accidents), the motorcyclist, and the severity of the accident—fatality—because we also had that information in the other databases. The RDB database extracted data about the direction, length, and size (radius, curvature) of horizontal elements, longitudinal gradients, and cross-sectional road dimensions. Since in both databases the data were related to road chainage, they could be merged (step 2).

Once the two main databases were merged, information about the rider’s direction of travel was obtained from the police data and media descriptions (step 3). This information was particularly important since it allowed us to point out how riders approached different horizontal elements. The earlier mentioned data, which were recorded in the TSA database, included the accident’s date, time, and location (including municipality). All this information was considered when searching for the accident in question to determine the (real) direction of travel of the motorcyclist. The data had to be identical to the additional data from the police, or if there was no data on the accident in the police database, we found the data in the online media according to the (identical) data mentioned above. Data were found for all road accidents. Regarding real direction, in the TSA database, we have information on the road number, road section and the road chainage at which the accident occurred. For the road section we also have information on the location (from location A to location B) in the direction of increasing road chainage. The data from the RDB database is also linked to the direction of increasing road chainage. Therefore, if the road accident (according to TSA data) happened on a certain section, which extents between place A and B, in the left curve (according to RDB database data), and from the description of the police/online media, we understood that the motorcyclist was driving in the direction from place B to place A: this means that the road accident happened in the right curve. The same approach was used for the longitudinal gradient data. If the motorcyclist’s direction of travel was opposite to the direction of increasing road chainage, a positive longitudinal gradient (according to the RDB data) meant that the actual road accident occurred in a negative longitudinal gradient.

Based on all this information, the initial merged database was modified (Figure 1—step 4), and the effective direction and dimension of the horizontal elements (especially curves) were added.

Table 1. Data recorded in the final, merged database.

Accident	Involved Users	External Conditions	Road	Road Elements
Severity	Age group	Statistical region	Name of the road	Curvature
Typology	Gender	Municipality	Road identification number	Radius
Cause	Nationality	Day	Name of a road section	Direction
			Road section identification number	Longitudinal slope
	Causer	Day time	Urban area/non-urban area	Pavement width
	Type of user	Weather	Pavement type
	Years of experience	Traffic	Road condition
	DUI *	Date	Location within the road
			Road chainage

* DUI: driving under the influence of alcohol or drugs.

summarizes all data considered in the final database.

Starting from the merged and modified database, a two-fold analysis was developed (Figure 2). First, a three-step general analysis of the obtained data was worked out: firstly, considering all fatal accidents with involved motorcyclists, then restricting them to the fatal accidents caused by a motorcyclist, and finally, deepening the analysis in single-vehicle accidents, i.e., accidents, which involved only the motorcycle rider. After that, a statistical analysis was worked out, and the correlation between factors leading to the accident and its happening was searched. To create the database and work out the initial analysis we used Excel, and SPSS to work out the correlation analysis. The first stage of analysis touched a total of 138 road accidents; the second stage reduced the number of accidents to 89—these are the accidents caused by motorcyclist, and the last one limited to 44 single-vehicle accidents of all the considered fatal road accidents.

4. Results

In this section, the results of all analyses are described. For clarity, results will be divided into sub-sections, which refer to each analysis stage described in Section 3. Additionally, the sub-sections related to stages 1–3 (Section 4.1, Section 4.2 and Section 4.3) are split into three parts, presenting the results associated with the involved road users, the cause and type of the accident and the design elements.

4.1. Analysis of All Fatal Road Accidents with Involved Motorcyclists

As shown in Figure 3, a total of 138 fatal road accidents involving motorcyclists happened in the 10-year period from 2013 to 2022. Their trend has been more or less stable, though years 2014 and 2021 stand out, with 37.7% and 44.9% more accidents than the mean number in the considered 10-year period. In addition, year 2022 is noticeable, with a decrease in road accidents to 7 in a year, i.e., 49.27% less than the period’s average. The majority of these accidents (90.58%) happened in non-urban area roads, i.e., on all two-lane roads, which run outside urban settlements (these are national primary and secondary roads, regional roads and local roads), while only the remaining 9.42% developed in urban areas. This finding could be expected since motorcyclists prefer riding on country, winding roads, and Slovenian non-urban area roads, which are particularly appealing for these road users.

4.1.1. Cause and Type of Road Accident

Two fundamental characteristics of road accidents—especially fatal ones—that should be clearly understood to improve the road safety of a specific group of road users are the reasons for the accident and the type of accident. The graph in Figure 4 shows that the majority of the fatal accidents under consideration were caused by unadjusted speed (65 road accidents, i.e., 47.10%), followed by three other factors, i.e., disrespect of yielding rules, incorrect driving direction and incorrect overtaking, which have much lower percentages—18.84% (26 road accidents), 15.22% (21 road accidents) and 12.32% (17 road accidents) respectively.

Figure 5 summarizes the results regarding the type of road accidents. It can be noticed that three kinds of road accidents have higher rates, i.e., head-on, side and object collisions. This is following previous results about the causes of the accidents: actually, it could be expected that an unadjusted speed could lead to the loss of control of the vehicle, causing the collision in an object or a head-on collision with another vehicle. Similarly, this can be said to be due to an incorrect overtaking or the partial use of the other driving lane. In addition, side collisions can be explained by the disrespect of yielding rules and probably also with some dynamics related to the other mentioned causes.

4.1.2. Involved Road Users

Interesting results are shown in Figure 6a,b. The first graph (Figure 6a) shows that in 89 road accidents (64.49%) of all considered accidents, the motorcycle rider caused the accident. That is in accordance with the causes leading to the accident that were mentioned in the previous section. In addition, it can be noticed (Figure 6b) that in the majority of cases (64 road accidents, i.e., 46.38%), two road users were involved in the accident, followed by single-vehicle accidents (44 road accidents, i.e., 31.88%).

When analyzing the driving experience (Figure 7)—in terms of years attaining the driving license—of the drivers causing the accidents, three age intervals stand out: drivers with 5 to 10 years of experience—18.12% (25 road accidents), novice drivers with less than 1 year of experience—15.94% (22 road accidents), and older drivers with 30 to 40 years of driving experience—14.49% (20 road accidents).

According to these results, the age of the riders causing the accident is between 18 and 34 years in 36.23% of the cases (50 road accidents) and over 45 in 47.82% (66 road accidents) of the accidents.

4.1.3. Design Elements

When analyzing the road geometrical characteristics, it can be noticed (Figure 8a,b) that the majority of the accidents involving motorcyclists happen on roads with good lateral and longitudinal conditions: as a matter of fact, 46.38% of the accidents happened on roads with a pavement width between 6 and 7 m (64 road accidents), and with 0% longitudinal gradient (70 road accidents). This can be explained by a lack of motorcyclists’ attention on these road segments, where the road conditions seem manageable to them. In addition, such conditions do affect the choice of a higher speed, which is relatable to the previous results about the causes leading to the accidents.

Focusing on the curve elements of the road, graphs in Figure 9a,b show that the majority of the accidents happened on left curves—51% (71 road accidents), followed by right curves—32% (44 road accidents) and only 17% (23 road accidents) of the considered accidents happened on straight road segments.

The radii of these curves are generally very large, ≥500 m, both for the left and right-handed curves. This could be explained by the behavior motorcyclists have to keep on curves, lean the vehicle and their body, and sometimes invade the lane in the opposite direction to be able to ride the curve (so-called “head-on” zone).

4.2. Analysis of Fatal Accidents Caused by Motorcyclists

In this section, the analysis focuses on 89 road accidents in which the motorcyclist was not only involved but also caused the accident. In addition, for this group of road accidents, the most tragic year was 2021—with 13 road accidents, and the vast majority—89.89% (80 road accidents)—happened on non-urban roads, confirming previous results. As highlighted in the following subsections, some patterns can be recognized in each analysis performed for this study.

4.2.1. Cause and Type of Road Accident

Once more, the most common cause of road accidents was unadjusted speed (Figure 10—63 road accidents). Nevertheless, it can be observed that this reason is followed by incorrect driving directions—14.61% (13 road accidents), and incorrect overtaking—12.36% (11 road accidents), instead of the disrespect of yielding rules, as it was highlighted in the previous analysis. This fact highlights an issue in the behavior motorcyclists keep, and that could be related to the infrastructural element they are riding on, e.g., the already mentioned lean angle they should keep on curves.

Interestingly, the majority of the accidents are a crash into an object or head-on collision, 31.46% (28 road accidents) and 26.97% (24 road accidents) of the road accidents, respectively, as can be noticed in Figure 11.

Of these accidents, in 46% of the cases (41 road accidents) the motorcyclist collided with a vehicle, while in 22% (20 road accidents) the motorcyclist veered off the road (run-off-road crashes).

4.2.2. Involved Road Users

When analyzing the patterns of road users involved in accidents caused by a motorcyclist, it is interesting to notice that the majority of these accidents involve 1 or 2 road users (49.44%—44 road accidents—and 30.34%—27 road accidents, respectively), highly decreasing when the number of participants increases (Figure 12). This also explains the interest of this study to focus on single-vehicle accidents.

The results related to the driving experience are slightly different from the previous analysis. As shown in Figure 13, the motorcyclists causing the most accidents (17 road accidents) are novice riders (with less than 1-year experience—19.10%). This rate is even higher than the rate of novice drivers causing a road accident (Figure 13). This group is followed by riders with 5 to 10 years of experience and riders with 30 to 40 years of experience.

The age of the motorcyclists also confirms these results: respectively in the age of 25–34, 45–54 and 55–64, as shown by the graph in Figure 14.

4.2.3. Design Elements

When analyzing the longitudinal and lateral characteristics of the road, results similar to those obtained in the previous section are obtained. As a matter of fact, the majority of the accidents happen on roads of appropriate width—50.56% of the accidents (45 road accidents) happen on a width between 6 and 7 m and with no longitudinal gradient—46.07% (41 road accidents).

Similarly, left curves are the most dangerous—51% of road accidents (45 road accidents), followed by right curves—38% (34 road accidents), both with high radiuses—more than 500 m. All these mentioned results are summarized in Figure 15.

4.3. Analysis of Single-Vehicle Accidents

This section develops the analysis of the 44 single-vehicle accidents that happened in the considered 10-year period. Similar to the previously considered groups of accidents, in this case, the trend of accidents through the years shows a fall in 2020—probably due to the COVID-19 restriction—and a peak in 2021. It is also confirmed that the great majority of single-vehicle accidents—81.82% (36 road accidents), happen in non-urban areas.

4.3.1. Design Elements

Also, for single-vehicle accidents, the most common cause leading to the accident is driving with unadjusted speed, followed by incorrect driving direction and incorrect overtaking (Figure 16a), resulting consequently in accidents involving the collision in an object (25 road accidents, i.e., 56.82% of the accidents, or in the vehicle overturn—31.82% (14 road accidents) of the cases (Figure 16b).

Deepening the analysis of single-vehicle accidents in an object, the graph in Figure 17 shows that in 43% of the cases (19 road accidents), the motorcyclist collides with an object on the roadside—which, unfortunately, is not precisely defined in the database, followed by the steel road-safety barrier—21% (9 road accidents), and a pole—14% (6 road accidents), demonstrating that road equipment should still be improved—from the point of view of positioning, material and shape, in order to reduce the consequences of such accidents.

4.3.2. Involved Road Users

Considering the characteristics of the involved road user—so of the only involved motorcyclist—it can be noticed that there is a slight difference with the previous results: indeed, single-vehicle accidents generally happened to riders with 5 to 20 years of experience (43.18% of the single-vehicle accidents, i.e., 19 road accidents) and belonging to a pretty extended age group, from 25 to 54 years old. A deeper analysis of the age group (Figure 18) revealed that 70% of the motorcyclists (7 road accidents) holding a driver’s license for 10 to 20 years are aged 35 to 44 years.

4.3.3. Design Elements

Similar to the previously analyzed road accidents, in this case, most of the accidents happened on roads wide enough (6–7 m) and without any particular longitudinal slope. Left curves are also the most problematic infrastructural elements, especially those with large radii (more than 500 m). These results are presented in Figure 19 and Figure 20.

4.4. Statistical Correlation Analysis

Previous sections have highlighted various factors characterizing fatal road accidents involving motorcyclists. Nevertheless, these patterns were not always the same, as the analysis deepened and focused on single-vehicle accidents. This part of the study aims to determine if a statistical correlation exists between any of the considered factors and the motorcyclist being the cause of the accident.

To have a statistically valid sample of data, all fatal accidents were considered—138, and an additional column with a binary indication of who was the causer was added. In this database, where the causer was the dependent variable, and all the other factors (see Table 1) were the independent variables, a Spearman rank order correlation (also known as Spearman’s rho) analysis was developed. This analysis was chosen because it has fewer limitations in comparison to Pearson’s correlation one, being able to analyze all monotonic relationships between continuous and ordinal variables, and it is much less sensitive to extreme values in comparison to Pearson’s test. The analysis was worked out first by referring to all variables belonging to human factors, then to the infrastructure, and finally to external conditions. The results obtained for each of these analyses are matrices containing the values of Spearman’s rho coefficient and the p-value for each couple of considered variables. A Spearman’s rho coefficient that is highly different from zero indicates a positive or negative correlation between the two considered variables. In contrast, a p-value lower than 0.05 indicates that the previously stated correlation is also statistically significant.

Table 2. Spearman’s correlation analysis between the causer and human factors.

		Driving Experience	Age	Gender
Causer	Spearman’s rho	0.793	0.730	0.981
	p-value	0.000	0.000	0.000

,

Table 3. Spearman’s correlation analysis between the causer and design elements.

		Driving Direction	Longitudinal Gradient	Pavement Width	Radius
Causer	Spearman’s rho	−0.073	−0.039	−0.229	−0.013
	p-value	0.397	0.654	0.007	0.88

and

Table 4. Spearman’s correlation analysis between the causer and external conditions.

	Causer
	Spearman’s rho	p-Value
Day	0.003	0.968
Day/night	0.073	0.393
Hour	−0.098	0.252
Weather	−0.065	0.448
Road condition	−0.053	0.540
Pavement type	−0.051	0.554
Traffic	0.038	0.657
Type of accident	0.601	0.000
Cause of the accident	−0.475	0.000
Urban/non-urban area	−0.032	0.710
Location within the road	−0.233	0.006
Category of the road	0.052	0.542

report the results of the analysis.

Results show that there are seven factors statistically significantly correlated with the causer of fatal accidents: three of them belong to the human factors and are—age, driving experience and gender; one is part of the infrastructure, and specifically its pavement width, while the last three belong to the external conditions, and are the type, cause and location of the accident on the road. In addition, it can be noticed that all these factors are positively correlated with the dependent variable, while only three of them (pavement width, causes, and road location) are negatively related to the causer.

The direction of the correlation parameter provides interesting information about the influence of each factor on the causer of the accidents. Starting from human factors, the positive correlation of driving experience and age indicates that more experienced drivers are more likely to be involved in fatal accidents, while the positive correlation of gender indicates that males are predominantly causers of these accidents. While the second statement could be more or less expected, due to the disproportion among male and female motorcyclists, the first finding is at first quite surprising. Nevertheless, two aspects should be considered: firstly, the accidents under analysis are fatal accidents, happened on very demanding roads. Secondly, often experienced drivers though being able to better estimate external conditions, they overestimate their capabilities, which could be one of reasons for their unadjusted behavior (see Figure 16a, the major cause for single-vehicle accident is unadjusted speed).

Focusing on the road geometry, a slightly negative correlation of the pavement width with the motorcyclist being the causer of the accidents, confirms that narrower roads are more dangerous for motorcyclists. This could be expected, since in narrower segments, especially curves, motorcyclists tend to lean with their body and vehicle towards the opposite lane, often invading it. Finally, the direction of the correlation parameter for the external conditions factors (Table 4) confirms the results found in the trend analysis. Additionally, some considerations should be made about the strength of the correlation, which is expressed by the magnitude of Spearman’s rho: the closer to 1 this parameter is, the stronger the correlation. It can be noticed that all three human factors have very high values of Spearman’s rho, underlining the strong correlation, therefore the strong impact these factors have on the motorcyclists causing the accident. Similarly, the type of accident has a high value as well. The other factors, though being statistically significant, have a weaker correlation, indicating that their lower influence on the motorcyclist as causer of the accident.

5. Discussion

The analyses worked out on the database of fatal road accidents with involved motorcyclists have highlighted some interesting common factors (

Table 5. Common factors of fatal road accidents with involved motorcyclist.

	All Fatal Accidents	Fatal Accidents with Motorcyclist Causer	Single-Vehicle Accidents
Cause	Unadjusted speed	Unadjusted speed	Unadjusted speed
Type	Head-on collision	Object collision	Object collision
Involved road users	Age, gender	Age, gender	Age, gender
Horizontal elements	Left curve, ≥500 m	Left curve, ≥500 m	Left curve, ≥500 m

) that can be useful in identifying measures to improve these road users’ safety.

Firstly, unadjusted speed and incorrect driving direction are the two most common causes leading to fatal accidents, especially, when the motorcyclist is causing them. These causes highlight a problem in the behavior of motorcyclists that should be tackled both in the early stages of the driving teaching activities and later with some additional training since—as it has been highlighted—many of the accidents have been caused by people with many years of riding experience. As a matter of fact, such behavioral mistakes could be due to overconfidence in their driving skills: though expert drivers should be able to better evaluate external conditions and the risk caused by it, they could over-estimate their capabilities, especially in terms of physical control of the vehicle. The most common types of road accidents are also tightly related to the highlighted causes, being in all the worked-out analyses of head-on accidents and object collisions. Head-on collisions can be explained by the motorcyclist’s incorrect driving direction, which generally tends to invade the opposite lane, especially when riding on curves. On the other side, object collision can be easily relatable to higher speed and the loss of control of the vehicle, suddenly riding off the road.

These hypotheses are also confirmed by the infrastructural elements leading to the majority of the accidents: if, on the one hand, segments with no longitudinal slope and enough pavement width have been stressed out, on the other one, left-handed curves are the most dangerous for motorcyclists. The study [19] highlighted curves as one of the most dangerous and demanding road sections—concerning motorcycle crashes and the severity of their consequences—given that 38% of crashes involving motorcycles occurred on curved road sections. This is also consistent with previous studies that indicated that horizontal curves are spots where the frequency of motorcyclist crashes is the highest [20] and where consequences are most severe [21].

The statistical significance of the results obtained by the road accident analyses has also been confirmed by the correlation analysis, which highlighted the driving experience, age, type and cause and road location among the most influencing factors for fatal accidents with involved motorcyclists. For example, study [22] also point out that old riders were more likely to die in a crash than young riders.

One way of preventing fatal accidents involving motorcyclists, or at least reducing their consequences, is to take additional infrastructural countermeasures. Two different types of countermeasures can be considered that can have a positive impact on motorcyclist safety: on the one hand, countermeasures that visually redirect motorcyclists and unconsciously influence their position on the road. An example of these countermeasures are horizontally painted areas that increase the centerline between the two lanes, keeping motorcyclists away from the road axis (preventing head-on) and consequently discouraging them from riding in the oncoming lane. On the other hand, measures relating to road equipment should also be considered, particularly with regard to its positioning, material and shape. Motorcyclist-friendly safety barriers, passive safety bollards, crash cushions for motorcyclists, and breakaway devices are just a few examples of things that can have a positive impact on the road safety of these road users.

6. Conclusions

Motorcyclist safety remains a major problem, both in the European Union and in other parts of the world. Although riding motorcycles has many advantages, there are also disadvantages, especially in terms of road safety. The biggest safety issue with motorcycles is related to their physical characteristics: they are single-track vehicles with low weight and powerful engines that achieve higher acceleration and a higher top speed than most other vehicles. In addition, motorcyclists are not protected by a safety shell, and their balance is highly dependent on the rider’s skills [19]. Furthermore, certain elements of road equipment that are perfectly safe for other road users can pose a major risk to motorcyclists. For all these reasons, it is necessary to deepen the knowledge of the factors that cause fatal accidents involving motorcyclists: this would make it possible to propose appropriate countermeasures that could reduce the number or at least the consequences of road accidents involving these road users.

The aim of the present study was to shed light on the safety level of motorcyclists in Slovenia by analyzing the latest traffic accident databases available at the national level in order to determine which factors influence the occurrence of fatal accidents involving motorcyclists. To conduct an accurate, up-to-date and comprehensive investigation, four complementary databases were merged and used as a starting point for this study. First, a three-stage general analysis of the data obtained was conducted: first, all fatal accidents involving motorcyclists were considered, then the data were limited to fatal accidents caused by a motorcyclist, and finally, the analysis was deepened to single-vehicle accidents. After that, a statistical analysis was developed to determine whether there is a correlation between the factors leading to the accident and the accident itself.

The results show that the vast majority of single-vehicle accidents (81.82%) occur outside urban areas, which is to be expected, as motorcyclists prefer roads with many curves and no intersections. The most common causes that lead to this type of accident are excessive speed, wrong driving direction and incorrect overtaking, which subsequently leads to accidents involving the collision in an object (56.82%) or the vehicle overturns (31.82%). In 43% of the single-vehicle accidents analyzed, the motorcyclist collided with an object of the roadside, followed by the steel safety barrier (21%), and a pole (14%). This demonstrates that roadside equipment should still be improved—in terms of positioning, material and shape—to reduce the consequences of such accidents.

Furthermore, the results of the correlation analysis show that there are seven factors statistically significantly correlated with the causer of fatal accidents: three of them belong to human factors, one is part of the road infrastructure, and the last three belong to external conditions.

Although the analysis developed gives an overview of the factors influencing fatal road accidents involving motorcyclists, it should be emphasized that it is limited to one severity level and that road accidents resulting in minor or serious injuries were not considered. The analysis of accidents involving motorcyclists with minor or serious injuries could also be of interest, especially if it leads to understanding the patterns associated with different levels of severity. This could also include the characterization of motorcyclists’ injuries in road accidents according to the Abbreviated Injury Scale (AIS). Likewise, it would be interesting to analyze the behavior of motorcyclists proactively, for example, by observing near misses on video recordings. This would make it possible to understand the process leading to near misses that could later turn into accidents, highlighting possible risky behaviors taken by the motorcyclists. Additionally, it should be also noticed that, due to the negligible percentage of motorcyclists driving on these roads in Slovenia, expressways and motorways were not considered in this study. Nevertheless, further works could consider also the behavior of motorcyclists on these roads, highlighting similarities and differences with their behavior on rural roads.

Finally, it would be interesting and useful to develop longitudinal studies aimed at evaluating the influence of infrastructural improvements and/or training measures on the number of fatalities, consequently assessing their impact on the improvement of road safety for motorcyclists.